• gareis <dhasenan gmail.com> May 26 2007
• gareis <dhasenan gmail.com> May 26 2007
• "Jarrett Billingsley" <kb3ctd2 yahoo.com> May 26 2007
• Deewiant <deewiant.doesnotlike.spam gmail.com> May 26 2007
• Don Clugston <dac nospam.com.au> May 29 2007
• "Jarrett Billingsley" <kb3ctd2 yahoo.com> May 29 2007
• Chris Nicholson-Sauls <ibisbasenji gmail.com> May 29 2007
• davidb <ta-nospam-zz gmx.at> May 26 2007

gareis <dhasenan gmail.com> writes:

---
import std.regexp;
int main () {
    char[] foo = "foo = bar";
    int i = std.regexp.find(foo, "[^\\]#");
    return 0;
}
---

This outputs "Error: invalid range". According to the documentation, this
should return -1, since the input does not contain an octalthorpe (#) that is
not preceded by a backslash.

Adding an octalthorpe doesn't help matters; it produces the same error. Also,
adding the attribute argument doesn't help.

With an example this small, I think this is probably a bug, but I'd like to
check here before submitting a report -- am I doing something wrong?

gareis <dhasenan gmail.com> writes:

Argh -- forget that. I forget that regexps require an extra level of escaping;
that should have been:
---
std.regexp.find(foo, "[^\\\\]#");
---

This is why I ask now rather than submitting bug reports first.

Thanks!

"Jarrett Billingsley" <kb3ctd2 yahoo.com> writes:

"gareis" <dhasenan gmail.com> wrote in message 
news:f399o6$n1f$1 digitalmars.com...
 Argh -- forget that. I forget that regexps require an extra level of 
 escaping;
 that should have been:
 ---
 std.regexp.find(foo, "[^\\\\]#");


If you use WYSIWYG strings, you can avoig the ugly double escaping:

std.regexp.find(foo, r"[^\\]#");

or

std.regexp.find(foo, `[^\\]#`);

I've also never heard of a pound sign being called an "octalthorpe"? 

Deewiant <deewiant.doesnotlike.spam gmail.com> writes:

Jarrett Billingsley wrote:
 I've also never heard of a pound sign being called an "octalthorpe"? 
 


http://en.wikipedia.org/wiki/Number_sign

-- 
Remove ".doesnotlike.spam" from the mail address.

Don Clugston <dac nospam.com.au> writes:

Deewiant wrote:
 Jarrett Billingsley wrote:
 I've also never heard of a pound sign being called an "octalthorpe"? 


 http://en.wikipedia.org/wiki/Number_sign
 


(BTW, what do you call ₤?)

"Jarrett Billingsley" <kb3ctd2 yahoo.com> writes:

"Don Clugston" <dac nospam.com.au> wrote in message 
news:f3grec$uqq$1 digitalmars.com...


 what do you call ??)


Oh I guess that's called a pound sign too, but whenever you say "pound sign" 
here it means almost certainly #.  We might call ? a "British pound sign" or 
so. 

Chris Nicholson-Sauls <ibisbasenji gmail.com> writes:

Jarrett Billingsley wrote:
 "Don Clugston" <dac nospam.com.au> wrote in message 
 news:f3grec$uqq$1 digitalmars.com...
 Where I come from, 'pound sign' has always meant ?, and # is 'hash'. (BTW, 
 what do you call ??)


 Oh I guess that's called a pound sign too, but whenever you say "pound sign" 
 here it means almost certainly #.  We might call ? a "British pound sign" or 
 so. 
 
 


I've oft heard it called a "Sterling".

-- Chris Nicholson-Sauls

davidb <ta-nospam-zz gmx.at> writes:

gareis schrieb:
 Argh -- forget that. I forget that regexps require an extra level of escaping;
 that should have been:
 ---
 std.regexp.find(foo, "[^\\\\]#");
 ---
 
 This is why I ask now rather than submitting bug reports first.
 
 Thanks!


---
int i = std.regexp.find(foo, r"[^\\]#");
---
works as well

david